PAC Confidence Sets for Deep Neural Networks via Calibrated Prediction

Sangdon Park, Osbert Bastani, Nikolai Matni, Insup Lee

Abstract: We propose an algorithm combining calibrated prediction and generalization bounds from learning theory to construct confidence sets for deep neural networks with PAC guarantees---i.e., the confidence set for a given input contains the true label with high probability. We demonstrate how our approach can be used to construct PAC confidence sets on ResNet for ImageNet, a visual object tracking model, and a dynamics model for the half-cheetah reinforcement learning problem.

Keywords: PAC, confidence sets, classification, regression, reinforcement learning

Code: https://github.com/sangdon/PAC-confidence-set

Community Implementations: [ 1 code implementation](https://www.catalyzex.com/paper/pac-confidence-sets-for-deep-neural-networks/code)